Response to the Editor -Basic Theme IV: Inductive and Deductive Reasoning

[Extract]
The Editor has raised a very timely and interesting debate on
inductive and deductive reasoning. However, I am slightly deviating
from this issue, and I will rather concentrate on a seminal paper by
Leo Breiman (2001) on “Statistical Modelling: The Two Cultures”
and later on I will try to link it to this topic. According to Breiman,
“There are two cultures in the use of statistical modelling to reach
conclusions from data. One assumes that the data are generated
by a given stochastic data model. The other uses algorithmic
models and treats the data mechanism as unknown”. In the first
case like regression models, logistic regression, and Cox-model,
the values of the parameters are estimated from the data and the
models are used for information and/or prediction. The second
case, which Breiman calls “The Algorithmic Modelling Culture”, the
analysis in this culture is complex and unknown. Their approach
is to find a function f(x) through an algorithm that operates on
X to predict Y and Breiman himself called this a black box. The
machine learning algorithms like Decision Tree, Random Forest,
and Stochastic Gradient Boosting and to some extent ANN falls
in this category.